Download Prophylactic Cranial Irradiation for Patients With Limited

Original Study
Prophylactic Cranial Irradiation for Patients With
Limited-Stage Small-Cell Lung Cancer With
Response to Chemoradiation
Patricia Tai,1 Avi Assouline,2 Kurian Joseph,3 Larry Stitt,4 Edward Yu5
Abstract
For patients with limited-stage small-cell lung cancer (SCLC) with an incomplete response (IR) to chemotherapy and chest radiotherapy, the benefit of prophylactic cranial irradiation (PCI) is not clear in the literature.
This report was based on 289 patients treated with curative intent, 93 of whom had incomplete response (IR).
These patients benefited from PCI, with a reduced rate of and a delayed time for the development of brain
metastases, although without significant overall or cause-specific survival (CSS) benefit. PCI could be considered for both complete and incomplete responders.
Background: Previous clinical studies have generally reported that prophylactic cranial irradiation (PCI) was given to
patients with a complete response (CR) to chemotherapy and chest radiotherapy in limited-stage small-cell lung
cancer (SCLC). It is not clear if those with incomplete response (IR) would benefit from PCI. Patients and Methods:
The Saskatchewan experience from 1981 through 2007 was reviewed. Patients were treated with chest radiotherapy
and chemotherapy with or without PCI (typical doses: 2500 cGy in 10 fractions over 2 weeks, 3000 cGy in 15 fractions
over 3 weeks, or 3000 cGy in 10 fractions over 2 weeks). Results: There were 289 patients treated for curative intent,
177/289 (61.2%) of whom received PCI. For the whole group of 289 patients, PCI resulted in significant overall survival
(OS) and cause-specific survival (CSS) benefit (P ⫽ .0011 and 0.0005, respectively). The time to symptoms of first
recurrence at any site with or without PCI was significantly different: 16.9 vs. 13.2 months (P ⫽ .0006). PCI
significantly delayed the time to symptoms of first recurrence in the brain: 20.7 vs. 10.6 months (P ⬍ .0001). The first
site of metastasis was the brain for 12.5% and 45.5% patients with CR with and without PCI, respectively (P ⫽ .02)
and in 6.1% and 27.6% of patients with IR with and without PCI, respectively (P ⫽ .05). For the 93 patients with IR,
PCI did not confer OS or CSS benefit (P ⫽ .32 and 0.39, respectively). Conclusions: Patients with IR benefited from
PCI, with a reduced rate of and a delayed time for the development of brain metastases, although without significant
OS or CSS benefit. PCI could be considered for all patients with limited-stage SCLC responding to chemoradiation.
Clinical Lung Cancer, Vol. 14, No. 1, 40-4 © 2013 Elsevier Inc. All rights reserved.
Keywords: Brain, Lung cancer, Metastasis, Radiotherapy, Survival
Introduction
The brain is a common site of metastasis for small-cell lung cancer
(SCLC). Unfortunately, because of the blood-brain barrier, chemother1
Department of Radiation Oncology, Allan Blair Cancer Center, Regina, Canada
Department of Radiation Oncology, Centre Clinique de la Porte de Saint Cloud,
Boulogne, France
3
Department of Radiation Oncology, Cross Cancer Institute, University of Alberta,
Alberta, Canada
4
Clinical Research Unit, London Regional Cancer Program, London, Canada
5
Department of Radiation Oncology, London Regional Cancer Program, University
of Western Ontario, London, Canada
2
Submitted: Mar 4, 2012; Revised: Apr 19, 2012; Accepted: Apr 23, 2012; Epub: June
5, 2012
Address for correspondence: Patricia Tai, MD, Radiation Oncology Department, Allan Blair
Cancer Center, Saskatchewan Cancer Agency, Regina, Saskatchewan S4T 7T1 Canada
Fax: 306-766-2845; e-mail contact: [email protected]
40
Clinical Lung Cancer January 2013
apy cannot achieve a sufficient tumoricidal dose. Some patients achieve
a complete response (CR) in the chest, only to have brain metastases
develop later. The literature shows that researchers have tested the efficacy of prophylactic cranial irradiation (PCI) in patients with limitedstage SCLC who achieved CR after combined chemotherapy and thoracic irradiation. They showed a trend toward improved survival.1-4
Rosen et al published a study in 1983 that suggested that any prolongation of survival would be restricted to patients with CR, because those
with residual extracranial cancer die of systemic metastases.1 Subsequent
to this, PCI was offered or recommended only to patients with CR in the
vast majority of clinical trials and guidelines.5-7 In the modern era, we
revisit the benefit of PCI in patients who have an incomplete response
(IR) to chemotherapy.
Limitations of past research are often due to the heterogeneous
nature of the studies: these include trials that mixed limited and
1525-7304/$ - see frontmatter © 2013 Elsevier Inc. All rights reserved.
http://dx.doi.org/10.1016/j.cllc.2012.04.005
Table 1 Overall Survival in Months for Patients With CR and IR
Variable
Median
Survival (mo)
Range (mo)
No. Alive at
1 Year
No. Alive at
2 Years
All Patients (N ⴝ 289)
19.7
4.0-248.2
225/289 (77.9%)
108/289 (37.3%)
Patients With CR (n ⴝ 185)
21.6
5.0-256.0
160/185 (86.5%)
83/185 (44.9%)
Patients With IR (n ⴝ 93)
14.3
4.0-102.6
79/93 (84.9%)
18/93 (23.7%)
Abbreviations: CR ⫽ complete response; IR ⫽ incomplete response.
After institutional review board approval, a chart review was undertaken for limited-stage SCLC in the Saskatchewan provincial database
from 1981 to 2007. We collected data on age, sex, performance status,
lactic dehydrogenase and hemoglobin levels, total number of chemotherapy cycles, radiotherapy dose for chest and PCI, timing of chest
radiotherapy and PCI, treatment interruptions, toxicities, development
of local recurrence, nodal recurrence, distant metastases in different organs, time to disease relapse, and survival status.
Patients had chest radiographs, bone scans, computed tomographic
scans, and other nuclear medicine scans if necessary for staging and
monitoring of response or recurrence during and after chemotherapy.
Patients were treated with systemic chemotherapy and chest radiotherapy. Typical dose fractionation of locoregional radiotherapy
to the chest was 4500 cGy in 25 fractions over 5 weeks, 5000 cGy in
25 fractions over 5 weeks, or 4000 cGy in 15 fractions over 3 weeks
(a commonly used Canadian regimen during the time in question).12
Typical PCI dose fractionation was 2500 cGy in 10 fractions over 2
weeks, 3000 cGy in 15 fractions over 3 weeks, or 3000 cGy in 10
fractions over 2 weeks. The decision to give PCI or not was made by
the treating oncologist after discussion with the patient and family.
For the purpose of this study, CR was defined as complete disappearance of all clinical evidence of tumor radiologically. PR was defined as
the decrease of the current sum of the products of maximum dimensions
of the measured lesions to 50% to 99% of the baseline sum. Minor
response (MR) was any decrease short of a 50% response. Hence, patients with IR could have PR or MR. Other response outcomes were
stable disease (SD), progressive disease (PD), and unknown.
The cause-specific survival (CSS) was the time interval from the date
of diagnosis to the date of death from SCLC, or the last follow-up date
for censoring purposes if the patient was alive and still being followed at
the time of analysis. The overall survival (OS) was defined as the time
from diagnosis to the date of death regardless of cause or the last follow-up date for censoring purposes if the patient was alive and still being
followed at the time of analysis. Progression-free interval (PFI) was cal-
O verall survival rate (%)
Patients and Methods
Figure 1 Overall Survival of the Whole Group of 289 Patients,
Divided According to Whether or Not They Received
PCI
100
90
80
70
60
50
40
30
20
10
0
P = .001
With PCI
No PCI
0
50
100
150
Time (months)
200
250
Figure 2 Cause-Specific Survival of the Whole Group of 289
Patients, Divided According to Whether or Not They
Received PCI
Cause-Specific Survival rate (%)
extensive stage together for analysis, patients treated with inadequate
doses of PCI, and PCI given only to patients with CR in some
studies,5-7 whereas in others PCI was given to both patients with CR
and patients with partial response (PR).8 Separate analysis of CR vs.
IR was seldom performed.9-11 Our group undertook this population-based study, which is unique to our knowledge, because it reports the patterns of recurrence in patients with CR and patients with
IR treated with and without PCI. We hope to answer the question of
whether PCI in patients with IR is worthwhile or not in the modern
era and describe the detailed outcome of patients with IR.
100
90
80
70
60
50
40
30
20
10
0
P = .0005
With PCI
No PCI
0
50
100
150
Time (months)
200
250
culated from the date of diagnosis to the date of symptomatic progression, as we would be more interested in the quality of life of patients
rather than asymptomatic radiologic disease progression. CSS, OS, and
PFI comparisons of patients who received PCI and patients who did not
were made using the Wilcoxon test13 to detect early separations in survival. Rates of brain metastases and brain recurrences between patients
Clinical Lung Cancer January 2013
41
Prophylactic Cranial Irradiation for Limited Stage SCLC
Table 2A Overall Survival Rates With and Without PCI
Variable
All Patients (N ⴝ 289)
Patients With CR (n ⴝ 185)
Patients With IR (n ⴝ 93)
PCI Status
1-Year OS (%)
2-Year OS (%)
PCI (n ⫽ 177)
84.9
42.6
No PCI (n ⫽ 112)
65.5
29.6
PCI (n ⫽ 132)
89.6
49.5
No PCI (n ⫽ 53)
77.8
33.7
PCI (n ⫽ 42)
69.7
18.1
No PCI (n ⫽ 51)
53.6
25.6
Wilcoxon Test P
.0011
.15
.32
Abbreviations: CR ⫽ complete response; IR ⫽ incomplete response; PCI ⫽ prophylactic cranial irradiation.
Table 2B CSS Rates With and Without PCI
Variable
All Patients (N ⴝ 289)
Patients With CR (n ⴝ 185)
Patients With IR (n ⴝ 93)
PCI Status
1-Year CSS (%)
2-Year CSS (%)
PCI (n ⫽ 177)
88.6
48.0
No PCI (n ⫽ 112)
75.5
32.3
PCI (n ⫽ 132)
92.5
55.4
No PCI (n ⫽ 53)
84.8
36.8
PCI (n ⫽ 42)
73.5
21.0
No PCI (n ⫽ 51)
59.3
28.3
Wilcoxon Test P
.0005
.10
.39
Abbreviations: CR ⫽ complete response; CSS ⫽ cause-specific survival; IR ⫽ incomplete response; PCI ⫽ prophylactic cranial irradiation.
Table 3A Rates of Brain Metastases As First Recurrence With and Without PCI in Deceased Patients
Response After
Chemoradiation
CR (n ⴝ 177)
IR (n ⴝ 88)
Treatment
Brain Recurrence
No Brain Recurrence
PCI (n ⫽ 128)
6
122
No PCI (n ⫽ 49)
5
44
PCI (n ⫽ 40)
2
38
No PCI (n ⫽ 48)
8
40
Fisher Exact Test P
.18
.10
Abbreviations: CR ⫽ complete response; IR ⫽ incomplete response; PCI ⫽ prophylactic cranial irradiation.
Table 3B Overall Rates of Brain Recurrence Before Death With and Without PCI in Deceased Patients
Response After
Chemoradiation
CR (n ⴝ 177)
IR (n ⴝ 88)
Treatment
Brain Recurrence
No Brain Recurrence
␹2 Test P
.002
PCI (n ⫽ 128)
24
104
No PCI (n ⫽ 49)
20
29
PCI (n ⫽ 40)
11
29
No PCI (n ⫽ 48)
15
33
.70
Abbreviations: CR ⫽ complete response; IR ⫽ incomplete response; PCI ⫽ prophylactic cranial irradiation.
who received PCI and those who did not were determined using ␹2
tests14 or when cell sizes were small, the Fisher exact 2-tailed test.
Results
From 1981 to 2007, there were 289 patients with limited-stage
SCLC treated with curative intent in the western Canadian province
of Saskatchewan. The median age was 65 years (range, 38-86 years).
42
Clinical Lung Cancer January 2013
The male-female ratio was 1.47 (178:121). Cisplatin-containing
chemotherapy was used in 67.5% (195/289) of patients. Eight patients did not have chemotherapy because of comorbidities or patient
choice. The response to chemoradiation was CR in 185 patients, PR
in 79 patients, MR in 14 patients, SD in 1 patient, PD in 3 patients,
and unknown in 7 patients. There were 177/289 (61.2%) patients
treated with PCI. Table 1 shows the OS for 185 patients with CR
Patricia Tai et al
Table 4 Progression-Free Interval With or Without PCI
Median (Range) Time to
First Symptomatic
Recurrence, Any Site
Variable
All Patients (n ⴝ 289)
Patients With CR (n ⴝ 185)
Patients With IR (n ⴝ 93)
PCI (n ⫽ 177)
16.9 (0.1-65.8)
No PCI (n ⫽ 112)
13.2 (1.6-57.9)
PCI (n ⫽ 132)
20.9 (5.4-65.8)
No PCI (n ⫽ 53)
13.1 (4.6-57.9)
PCI (n ⫽ 42)
12.2 (3.4-35.7)
No PCI (n ⫽ 51)
9.4 (1.8-37.1)
Wilcoxon
Test P
Median (Range) Time to
First Brain Metastatic
Symptoms
.0006
.12
.39
20.7 (8.4-40.7)
10.6 (1.8-24.6)
21.1 (8.4-40.7)
12.4 (4.1-21.0)
18.5 (9.5-34.5)
8.9 (1.8-14.6)
Wilcoxon
Test P
⬍ .0001
⬍ .0001
.03
Abbreviations: CR ⫽ complete response; IR ⫽ incomplete response; PCI ⫽ prophylactic cranial irradiation.
and 93 patients with IR. Figures 1 and 2 and Table 2 show that for
the whole group of 289 patients, PCI was associated with significant
OS and CSS benefit (P ⫽ .0011 and 0.0005, respectively). In the
subgroup analysis among the 185 patients with CR, PCI appeared to
confer better OS and CSS, although it was not statistically significant
(P ⫽ .15 and 0.10, respectively). For the 93 patients with IR, PCI did
not improve OS or CSS (P ⫽ .32 and 0.39, respectively). For the
subgroup of 62 patients with IR who received cisplatin chemotherapy, PCI did not improve OS or CSS (P ⫽ .84 and 0.98,
respectively).
There were 48 patients with CR in whom disease recurred after
PCI, 6 of whom had brain as the first site of recurrence, ie, 12.5% of
the first recurrence was in the brain despite PCI. For patients with
CR without PCI, 45.5% (5/11) had a first recurrence in the brain.
The difference is statistically significant (P ⫽ .02). Among patients
with IR, first recurrence in the brain was 6.1% (2/33) with PCI vs.
27.6% (8/29) without PCI (P ⫽ .051).
Table 3A shows rates of brain metastases as first recurrence,
with and without PCI, in the 274 patients who had died. There is
a trend for PCI to reduce the incidence rate of brain metastases as
the first recurrence site in patients with IR (P ⫽ .10). Table 3B
shows the incidence rates of disease recurrence in patients in
whom there was brain recurrence within their lifetimes with and
without PCI.
Table 4 shows the time to symptoms of first recurrence at any
site and in the brain, with or without PCI. For the whole group of
289 patients, the times to symptoms of first recurrence at any site,
with or without PCI were significantly different: 16.9 vs. 13.2
months (P ⫽ .0006). PCI significantly delayed the time to symptoms of first recurrence in the brain: 20.7 vs. 10.6 months (P ⬍
.0001). The time to development of the first brain recurrence was
almost twice as long for those who received PCI compared with
those who did not (P ⬍ .0001 and 0.03, respectively for patients
with CR and those with IR).
Discussion
There had been no national consensus on PCI until 2000, when
Drs Jaro Kotalik and Edward Yu wrote the practice guideline for
Ontario.15 The National Cancer Comprehesive Network guideline
recommended PCI for patients with 90% to 100% response in the
past. The most current guideline, published in 2011, recommends it
for all patients with CR and PR—ie, not stable or progressive dis-
ease.16 Our current study gives support for the use of PCI in patients
with IR. There is a trend for PCI to reduce the incidence rate of
brain metastases as the first recurrence site (P ⫽ .10). The symptom-free time for brain metastases is significantly longer after PCI
(P ⫽ .03). Brain metastases adversely affect the quality of life for
the patient, so PCI is justified even if there is no survival benefit in
patients with IR.
Our results are consistent with the literature: the reported median
survival after a CR was 18.2 months, with a median survival of 9.9
months for those showing a PR.17 In another series, the median
survival for the complete responders was 11.7 months, whereas the
partial responders survived for a median of 9.7 months.18 The
European Organisation for Treatment and Research of Cancer study
on extensive SCLC showed a benefit for PCI for patients with any
response.19 More oncologists are applying the same argument on
limited-stage disease as well.
Strengths of this study include relatively large patient numbers,
analysis of patients with CR and those with IR separately, documented times of first symptomatic recurrence, and consecutive patients from population data reflecting real-world community practice without the enrollment bias of a clinical trial. Weaknesses
include the retrospective nature of the study, possible bias with more
motivated and fit patients among the PCI group, current better imaging and treatments, and the reduced sample size when looking
within CR and IR subgroups. Therefore the study is limited to detect
differences between subgroups.
Despite the retrospective nature of this study, it is useful to
analyze the value of PCI in patients with CR and those with IR
separately. Randomized studies would not be possible in view of
the small number of cases of limited-stage SCLC. They also have
enrollment bias, with more motivated, fit patients in urban areas
joining them.
The main focus of this study is the additive value of PCI in patients
who had different degrees of response to chemoradiation because
chemotherapy by itself is not effective for preventing brain metastases
due to the blood-brain barrier. Therefore the actual chemotherapeutic agents used should not affect our result.
Conclusion
The current study shows that in the overall group of 289 patients
with limited-stage SCLC, PCI confers OS and CSS benefit (P ⫽
.0011 and 0.0005, respectively). Although PCI did not significantly
Clinical Lung Cancer January 2013
43
Prophylactic Cranial Irradiation for Limited Stage SCLC
improve survival in the IR subgroup, brain metastasis with its attendant
neurologic complications was effectively prevented or delayed.
Our finding has significant impact on daily clinical practice. First,
many questions surround the definition of IR concerning what investigations should be performed and when the response should be
measured. With modern imaging and chemotherapy, the measurement of response to chemoradiation is more sophisticated. Even with
modern imaging, response measurement is still challenging at times.
Residual radiologic abnormalities may be due to fibrosis/collapse/
consolidation rather than residual tumor. Second, measuring the
response right after chemoradiation may not allow the tumor to
shrink completely. Omitting PCI if residual disease is present can
end up causing a delay in administering PCI, allowing brain metastases to develop while waiting for CR to occur. This study provides
the justification to give PCI to patients with IR.
Because it is difficult to differentiate CR from IR accurately
despite modern imaging, and the lung primary tumor may continue to shrink after chemoradiation, PCI could be considered for
all patients with limited-stage SCLC who respond to
chemoradiation.
Clinical Practice Points
●
●
●
It is commonly believed that any potential benefit of PCI would be
restricted to patients with CR because those with residual extracranial cancer die of systemic metastases. Incomplete responders also
have a higher chance of local recurrence.
In the modern era, new findings in our current study give support
for the use of PCI in patients with IR. There is a trend for PCI to
reduce the incidence rate of brain metastases as the first recurrence
site (P ⫽ .10). The symptom-free time for brain metastases is
significantly longer after PCI (18.5 vs. 8.9 months; P ⫽ .03).
Since brain metastases adversely affect the patient’s quality of life,
PCI is justified even if there is no survival benefit for patients with
IR. Because it is difficult to differentiate CR from IR accurately
despite modern imaging, PCI could be considered for all patients
with limited-stage SCLC who respond to chemoradiation.
Acknowledgments
We thank Saskatchewan Cancer Agency, who provided research
grants, and all radiation and medical oncologists in both Allan Blair
and Saskatoon Cancer Centers who treated and followed lung cancer
patients over the study period.
44
Clinical Lung Cancer January 2013
Disclosure
The authors have stated that they have no conflicts of interest.
References
1. Rosen ST, Makuch RW, Lichter AS, et al. Role of prophylactic cranial irradiation in
prevention of central nervous system metastases in small cell lung cancer. Potential
benefit restricted to patients with complete response. Am J Med 1983; 74:615-24.
2. Gregor A. Prophylactic cranial irradiation in small cell lung cancer (SCLC) makes a
comeback. Clin Oncol (R Coll Radiol) 1997; 9:148-9.
3. Arriagada R, Pignon JP, Laplanche A, et al. Prophylactic cranial irradiation for
small-cell lung cancer. Lancet 1997; 11:349:138.
4. Niiranen A, Holsti P, Salmo M. Treatment of small cell lung cancer. Two-drug
versus four-drug chemotherapy and loco-regional irradiation with or without prophylactic cranial irradiation. Acta Oncol 1989; 28:501-5.
5. Arriagada R, Le Chevalier T, Borie F, et al. Prophylactic cranial irradiation for patients with
small-cell lung cancer in complete remission. J Natl Cancer Inst 1995; 87:183-90.
6. Bremnes RM, Sundstrøm S, Vilsvik J, et al. Multicenter phase II trial of paclitaxel,
cisplatin, and etoposide with concurrent radiation for limited-stage small-cell lung
cancerMulticenter phase II trial of paclitaxel, cisplatin, and etoposide with concurrent radiation for limited-stage small-cell lung cancer. J Clin Oncol 2001; 19:
3532-8.
7. Woo IS, Park YS, Kwon SH, et al. A phase II study of vp-16-fosfamide-cisplatin
combination chemotherapy plus early concurrent thoracic irradiation for previously untreated limited small cell lung cancer. Jpn J Clin Oncol 2000; 30:
542-6.
8. Bogart JA, Herndon JE 2nd, Lyss AP, et al. 70 Gy thoracic radiotherapy is feasible
concurrent with chemotherapy for limited-stage small-cell lung cancer: analysis of
cancer and Leukemia Group B study 39808. Int J Radiat Oncol Biol Phys 2004;
59:460-8.
9. Murray N, Sheehan F. Limited stage small cell lung cancer. Curr Treat Options
Oncol 2001; 2:63-70.
10. Ardizzoni A, Fusco V, Pennucci C, et al. Alternating chemotherapy with cyclophosphamide, doxorubicin, vincristine and cisplatin, etoposide followed by prophylactic
cranial and thoracic irradiation for small cell lung cancer (SCLC): long-term results.
Anticancer Res 1991; 11:681-4.
11. Aroney RS, Aisner J, Wesley MN, et al. Value of prophylactic cranial irradiation
given at complete remission in small cell lung carcinoma. Cancer Treat Rep 1983;
67:675-82.
12. Murray N, Coy P, Pater JL, et al. Importance of timing for thoracic irradiation
in the combined modality treatment of limited-stage small-cell lung cancer. The
National Cancer Institute of Canada Clinical Trials Group. J Clin Oncol 1993;
11:336-44.
13. Wilcoxon F. Individual comparison by ranking methods. Biometr Bull 1945;
1:80-3.
14. Analyze a 2 ⫻ 2 contigency table. GraphicPad software. Available at: http://
www.graphpad.com/quickcalcs/contingency1.cfm. Accessed: May 24, 2012.
15. Kotalik J, Yu E, Markman BR, et al. Practice guideline on prophylactic cranial
irradiation in small-cell lung cancer. Int J Radiat Oncol Biol Phys 2001; 50:309-16.
16. National Cancer Comprehensive Network. Available at: http://www.nccn.org/
professionals/physician_gls/pdf/sclc.pdf. Accessed: May 24, 2012.
17. Bishop JF, Kefford R, Raghavan D, et al. Etoposide, carboplatin, cyclophosphamide
and vincristine in previously untreated patients with small-cell lung cancer. Cancer
Chemother Pharmacol 1990; 25:367-70.
18. Broder LE, Selawry OS, Charyulu KN, et al. A controlled clinical trial testing two
potentially non-cross-resistant chemotherapeutic regimens in small-cell carcinoma
of the lung. Chest 1981; 79:327-35.
19. Slotman B, Faivre-Finn C, Kramer G, et al. Prophylactic cranial irradiation in
extensive small-cell lung cancer. N Engl J Med 2007; 357:664-72.